Evolution of Bentonite Under High Temperature Heating and Hydration: Bench-Scale Laboratory Experiments and Coupled Thermo-Hydro-Mechanical Modeling

: Bentonite buffer in the geological repository for high-level radioactive waste undergoes the heating from the waste package and hydration from the geological formation and goes through coupled thermo-hydro-mechanical-chemical (THMC) changes over the life span of a repository. For better understanding of such process under higher temperatures we report bench-scale laboratory experiments with heating up to 200°C and the corresponding THM model. The bench-scale laboratory experiments included two test columns, with the non-heated control column undergoing only hydration, and a heated column experiencing both heating in the center up to 200°C and hydration from a sand-clay boundary surrounding the column. During the experiment, we took frequent X-ray CT images to provide insight into the spatio-temporal evolution of THMC due to heating, hydration, bentonite swelling/compression. Based on the experiment setup, 2-D axisymmetric simulations were performed for the heated column and the mechanical changes were investigated in 3-D. The model first matched the temperature evolution with step-wise temperature boundary conditions at the heater and calibrated the thermal conductivity and specific heat of the materials. Then model interpreted the spatio-temporal distribution of bulk density by considering the combined effect of hydration, fluid pressure, and porosity change due to swelling/compression. 1 INTRODUCTION Engineered barrier system (EBS) with bentonite buffer has been proposed in the repository design for spent nuclear fuel and waste because of its low permeability, high swelling capacity and radionuclide retention, as well as thermal stability among other desired characteristics. After the emplacement of waste canisters and backfill materials, bentonite will be simultaneously heated from the decaying radioactive waste and hydrated from the surrounding host rock. These perturbations trigger coupled THMC (thermal-hydrological-mechanical-chemical) processes, involving (1) moisture transport controlled by multiphase flow and large thermal gradient near the heat source; (2) swelling and shrinkage due to bentonite hydration or de-hydration; (3) dilution/concentration, migration and exchange of ions impacted by moisture/thermal interactions, (4) dissolution/precipitation and mineral phase transformation, etc. (Zheng et al., 2010). While bentonite behavior for temperature <100°C has been well-documented based on laboratory and field experiments, data is very limited for high temperature (>100°C) conditions (Zheng et al., 2017). Itis important to understand the behavior of bentonite at high temperature for the disposal of high-level nuclear waste to expand the knowledge base regarding the perturbation of bentonite buffer and open the possibility of the designing repository with higher thermal limit. The performance assessment of repository requires prediction of the long-term THMC evolution of bentonite. The reliability of model prediction is hinged on reliable constitutive laws and parameterization to describe the THMC process at scale appropriate for model parameterization for large scale model. Models for the large scale in situ test, e.g. Zheng et al. (2020b), showed that parameters calibrated from modeling column tests can be used for large scale model. Currently, a high temperature experiment in a crystalline rock environment, called HotBENT, is being conducted under the leadership of NAGRA with several international partners (García-Siñeriz and Tuñón, 2020). This full-scale, high-temperature experiment will be conducted at the Grimsel Test Site. Such large-scale tests are extremely important for better understanding of the bentonite EBS system behavior under high temperature and conditions with strong thermal, hydraulic and density gradients. The main objective of the HotBENT experiment is to evaluate whether higher repository temperature would trigger mechanisms that compromise the barrier functions of the engineered system and the host rock. To complement the field scale HotBENT experiment, we had conducted a benchtop-scale laboratory experiment to obtain well-characterized datasets for understanding bentonite THMC processes under heating and hydration for model parameterization and benchmarking (Chang et al., 2021). In this paper, coupled THM models were developed to interpret the data collected from the column test, attempting to calibrate key constitutive relationships and parameters. In the following sections, we will first briefly describe the columns tests and then discuss the THM models